Is Ubiquitous A.I. Writing Actually Inevitable?

Max Read’s recent breakdown of his own workflow lands less like a prediction and more like a confession. He isn’t talking about a hypothetical future where robots write novels. He is describing a Tuesday morning. His AI tool is hooked into his Gmail, his Google Calendar, his Granola AI transcription service, and his Notion notes. He has even engineered a custom skill for Claude to replicate the specific cadence of his writing, complete with a set of “10 commandments” to mimic the style of Alex Heath.

Read isn’t predicting the future; he is living it. And for anyone wondering if the flood of AI-generated text is a temporary wave or the new ocean floor, Read’s setup offers a chilling answer: it is becoming the infrastructure itself.

The debate over adoption is over. We aren't debating whether we will use these tools anymore. The real friction point is what happens when the tool becomes the environment.

The Invisible Infrastructure

There is a persistent belief that AI writing is a distinct layer we can toggle on or off, like a filter in Photoshop. That assumption is failing. The technology is following Weiser’s vision of pervasive, invisible computing. It is not a destination; it is the road.

When Read connects his calendar and email to an LLM, the AI stops being a separate app he visits. It becomes an extension of his cognitive load. It knows his schedule, his inbox, and his past notes. This integration removes the friction of context switching. You don’t open a document and ask it to “write a summary of my week.” The system already has the data. You don’t copy-paste meeting transcripts into a prompt box; the AI ingests them in the background.

This is where the comparison to high-volume blogs or Facebook fake news becomes relevant. Those phenomena were inevitable not because people wanted them, but because the underlying platforms lowered the cost of distribution to near zero. AI writing does the same for creation. When the barrier to drafting drops from hours to seconds, volume explodes. Publishers are already drowning in content because the cost of entry has vanished.

For knowledge workers, this means the central repository matters. Read uses Notion as the brain for this operation. It stores the “10 commandments,” the past articles, and the instructions on how his newsletters should sound. If you aren’t organizing your context this way, you aren’t just missing a feature; you are missing the leverage point where human intent meets machine execution. But this convenience comes with a quiet price: the boundary between your thought and the machine’s suggestion blurs until you can no longer tell which came first. You start thinking in prompts instead of sentences.

The Cost of Convenience

If the infrastructure is inevitable, is the output? Optimism hits a wall when you look at the results. We are already seeing the early warning signs of homogenization.

Ubiquitous streaming subscriptions and algorithmic feeds have spent the last decade flattening creativity into an infinite scroll of mediocre, repetitive styles. The goal was engagement, and the result was a risk-averse culture. AI writing, trained on the very output of that culture, risks cementing this ceiling. The models learn from the average, and the average is rarely interesting.

Read’s custom skill is a fascinating defense against this. By explicitly instructing the AI to write like Alex Heath, he is fighting the model’s default tendency to write like everyone else. The default setting of these models is to be agreeable, safe, and statistically probable. That sounds nice until you realize it means your writing will start sounding like the average of everything the model has ever read. It favors passive voice, hedging language, and generic transitions.

This is the trade-off. You gain speed, but you lose the friction that often forces distinct voice. When an AI can draft a newsletter in the time it takes to brew coffee, the incentive structure shifts. Why spend three days refining a metaphor when the machine can offer three options in ten seconds? The danger isn’t that AI will write badly. It is that it will write competently, and competent is the enemy of great.

We see this in the education sector, where teachers are scrambling to detect AI-generated text. They assume AI lacks errors or follows a specific pattern. But as the models adapt, the detection game becomes an arms race. The real issue isn’t cheating; it’s atrophy. If students (or professionals) outsource the struggle of drafting, they lose the muscle memory required to construct complex arguments. They forget how to find their own rhythm because the machine provides one for them.

The Security and Dependency Trap

There is a concrete downside to this level of integration that goes beyond style. When you connect your Gmail and Calendar to a third-party AI service, you are handing over the keys to your professional life.

Read’s workflow requires the AI to read emails and access calendar events to generate context. This creates a massive attack surface. A breach in that AI tool doesn’t just leak your documents; it leaks your schedule, your contacts, and your private correspondence. We are trading privacy for productivity. If the vendor suffers a data breach, your entire professional history is exposed to the highest bidder.

Furthermore, there is the issue of lock-in. If you build your entire writing style and knowledge base inside a specific ecosystem—custom skills, specific prompts, proprietary databases—migrating away becomes difficult. You become dependent on the model that knows your voice best. If that model changes its pricing, shuts down, or alters its terms, your workflow collapses.

This isn’t a hypothetical risk. We have seen how quickly SaaS tools can pivot or disappear. Building your business logic on top of a foundation that changes its terms weekly is a gamble. The “inevitability” of AI writing assumes the tools remain stable and accessible. They might not. You might find yourself locked into a platform that no longer supports the specific API endpoints your custom skills rely on.

The Human Editor is the New Bottleneck

Inevitability doesn't mean surrender. Ubiquitous AI writing is inevitable in the sense that the technology will be everywhere. It will be embedded in your email client, your word processor, and your project management dashboard. But ubiquitous does not mean dominant.

The bottleneck is shifting. It is no longer about generating words. It is about curating them. The value of the writer is no longer in the first draft; it is in the edit. The ability to recognize the "slop," to spot the hallucinations, to inject the specific nuance that the model smoothed over—that is where the human remains essential.

Read’s setup works because he is the editor-in-chief of his own AI. He didn’t just let the model write; he built a skill to constrain it. He provided the guardrails. Without that human oversight, the output drifts toward the mean. The human job is now to cut the fluff, verify the facts, and ensure the tone matches the brand.

We are moving toward a world where the ability to command an AI is a basic literacy, but the ability to judge its output is a premium skill. The tools will do the heavy lifting, but they cannot make the call on whether the lift matters. They cannot decide if a story is worth telling.

Bottom Line

Ubiquitous AI writing is inevitable because the economic incentives align perfectly for its adoption. It reduces cost, increases speed, and integrates into existing workflows. However, the quality of that writing is not inevitable. It requires active management, strict guardrails, and a willingness to reject the easy answer.

If you adopt this technology, do not treat it as a replacement for your brain. Treat it as a junior associate that reads everything you write and learns from your mistakes. Connect it to your Notion notes, yes, but keep your critical thinking offline. The future of writing isn’t about who can generate the most text; it’s about who can cut through the noise to say something true. The machine can give you the draft, but only you can give it the soul.

Sources: https://maxread.substack.com/p/is-ubiquitous-ai-writing-inevitable | https://case.edu/weatherhead/xlab/about/news/ai-eating-world-why-ubiquitous-intelligence-inevitable-and-how-it-will-happen | https://www.thriveducation.org/2025/07/01/ai-is-ubiquitous-where-do-we-start/ | https://www.sarrkiverse.com/2025/09/ai-future-how-ai-will-shape-our-world.html